How does Genesys score memories differently from vector stores?

Genesys uses a multiplicative formula: decay_score = relevance × connectivity × reactivation. A memory must score on all three axes to survive—so a highly connected memory that is never recalled will still decay, and a frequently accessed memory that is causally orphaned will still fade. This prevents memories from accumulating indefinitely just because they exist.

How did Genesys perform against competitors on the LoCoMo benchmark?

Genesys scored 85.55 ± 0.37 (certified over 10 runs in July 2026, with a frozen setup of gpt-4o-mini answerer and judge at temperature 0 across 1,540 test cases and 4 categories), compared with Zep at 75.14 and Mem0 at 66.9 on comparable published setups. The oracle retrieval ceiling under this protocol is 94.9.

Does the open-source version require a database?

No. The core library (genesys-memory) is an in-memory causal graph with optional JSON persistence via GENESYS_PERSIST_PATH and has zero database dependencies. Additional backends (Postgres, FalkorDB, MongoDB, Obsidian) and REST API are part of the separate hosted product at genesys-api.astrixlabs.ai.

What platforms can Genesys integrate with?

The system speaks MCP natively and can be connected to Claude, ChatGPT, and any MCP-compatible app. It ships as a stdio MCP server (python -m genesys_memory) that can be added to Claude Desktop or Claude Code.

Large Language ModelsOpen-Source AIHacker NewsJul 21, 2026Jul 22, 2026, 01:01 JST

Genesys: open-source causal-graph memory for AI agents launches

Send on LINE

3 Key Points

What happened
Astrix Labs released Genesys, an open-source memory system for AI agents that scores memories using a multiplicative formula (relevance × connectivity × reactivation), organizes them in a causal graph, and actively forgets irrelevant ones. The system scored 85.55 ± 0.37 on the LoCoMo benchmark (certified over 10 runs in July 2026), outperforming Zep (75.14) and Mem0 (66.9) on comparable setups. The core library has zero database dependencies and ships with optional JSON persistence; a hosted version with Postgres and REST API is available separately.
Why it matters
Most AI memory systems dump everything into vector stores, giving recall without understanding—the 500th memory buries the ones that matter. Genesys addresses this by combining relevance decay, causal connectivity, and reactivation frequency so a memory must score on all three axes to survive, and by actively pruning stale, orphaned memories rather than accumulating them forever. This matters for any AI agent (in ChatGPT, Claude, or custom MCP apps) that needs to learn and reason about causality rather than just retrieve similar text.
What to watch
The open-source library is available now at github.com/Astrix-Labs/genesys with pip install genesys-memory; it integrates with Claude, ChatGPT, and any MCP-compatible app via a stdio server. The hosted product at genesys-api.astrixlabs.ai offers additional storage backends (Postgres, FalkorDB, MongoDB, Obsidian), multi-user auth, and a REST API. Benchmark reproduction scripts are included in the repo so users can verify the LoCoMo results themselves.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Get news like this every morning, free

Context & Analysis

Genesys addresses a fundamental problem in AI memory: vector-store systems treat all memories equally, ranking them only by semantic similarity, which causes important early context to disappear under a flood of later entries. By introducing a causal graph and a multiplicative decay formula, Genesys ensures that memories must remain relevant, well-connected to other memories, and actively recalled to survive. This mimics how human memory works—facts fade unless reinforced or integrated into a causal narrative.

The benchmark results are self-reported and were conducted with Genesys's own harness (excluding category 5, adversarial questions with disputed ground truth), so they are directional rather than independently verified. However, the authors have published full reproduction scripts and methodology, allowing users to run their own evaluations. The 85.55 score on LoCoMo, while higher than Zep (75.14) and Mem0 (66.9), sits below the oracle ceiling of 94.9, suggesting room for improvement and a realistic assessment rather than overstatement.

The open-source strategy is significant: by shipping a zero-dependency core library with optional JSON persistence, Genesys lowers the barrier to adoption for developers who want to experiment without cloud infrastructure, while offering a hosted version with Postgres, multi-user auth, and REST API for production use. Integration with MCP (Model Context Protocol) means the system can work with multiple AI platforms—Claude, ChatGPT, and any future MCP-compatible tool—rather than being locked to a single vendor.

FAQ

How does Genesys score memories differently from vector stores?: Genesys uses a multiplicative formula: decay_score = relevance × connectivity × reactivation. A memory must score on all three axes to survive—so a highly connected memory that is never recalled will still decay, and a frequently accessed memory that is causally orphaned will still fade. This prevents memories from accumulating indefinitely just because they exist.
How did Genesys perform against competitors on the LoCoMo benchmark?: Genesys scored 85.55 ± 0.37 (certified over 10 runs in July 2026, with a frozen setup of gpt-4o-mini answerer and judge at temperature 0 across 1,540 test cases and 4 categories), compared with Zep at 75.14 and Mem0 at 66.9 on comparable published setups. The oracle retrieval ceiling under this protocol is 94.9.
Does the open-source version require a database?: No. The core library (genesys-memory) is an in-memory causal graph with optional JSON persistence via GENESYS_PERSIST_PATH and has zero database dependencies. Additional backends (Postgres, FalkorDB, MongoDB, Obsidian) and REST API are part of the separate hosted product at genesys-api.astrixlabs.ai.
What platforms can Genesys integrate with?: The system speaks MCP natively and can be connected to Claude, ChatGPT, and any MCP-compatible app. It ships as a stdio MCP server (python -m genesys_memory) that can be added to Claude Desktop or Claude Code.

Hacker NewsRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.